Note : Les descriptions sont présentées dans la langue officielle dans laquelle elles ont été soumises.
. . ", .
ALLIED COLLOIDS LIMITED 60/2273/01
SYNTHESIS OF VINYL ESTERS
It is known to be able to make esters of carboxylic
acids by an ester interchanye reaction according to the
reaction scheme
RCOOR + R OH = RCOOR + R OH
In this typically R is an aliphatic or aromatic group, Rl
is a methyl or ethyl group and R2 is an aliphatic group
containing more carbon atoms than R1. A catalyst is
generally used in order to promote the desired exchange
reaction. Various compounds have been proposed for use
as the catalyst. For instance in British Patent
Specification 1573071 it is proposed to use a chelate of
a beta-diketone with zirconium or calcium and the use of
a mixture of such chelates is exemplified. Usually
however the catalyst is a metal alcoholate and it has
been proposed to form the alcoholate from the alcohol
that is to be reacted in the ester interchange ~eaction.
When the ester is unsaturated the prior art processes
often lead to the formation of impurities and may involve
difficult separation procedures.
In our Canadian Patent Application No.
439,608 we review the relevant prior art
associated with such ester interchange reactions and
describe a process for the production of vinyl esters
that is conducted in the presence of a metal alcoholate
and in which the metal alcoholate provides the groups
that go into the desired ester, the reaction being
conducted in the absence of free alcohol. In particular
in that process we make an ester of the formula R3CooR4
where R3 is CH2=CH- or CH~=CICH3)- and R4 contains at
least four carbon atoms and is selected from alkyl,
cycloalkyl and aminoalkyl by reaction of a compound of
2~a -
~ ; c
- 2 -
the formula R3CooRl, where Rl is Cl 3 alkyl, in the
presence of a metal alcoholate formed from an a~cohol
R40H and in this process the metal alcoholate provides
the groups R4 that are utilised in forming the desixed
ester and reacts with the compound R3CooR1, the reaction
is conducted in the substantial absence of water or
reactive alcohol and the metal is selected from titanium,
aluminium, zirconium, calcium and magnesium. We
mentioned that mixtures cf metals may be used and
exemplify a process using a mixture of titanium and
zirconium.
Processes in which the metal of the alcoholate is
magnesium or calcium can be operated at relatively low
temperatures, for instance 10 to 50C, to give good
yields of the desired product. However these metal
alcoholates are solids that are insoluble in the reaction
mixture and in practice they can tend to result in the
formation of a semi-colloidal mixture having physical
characteristics such that separation, further reaction
and recycling of the metal alcoholates, as described in
Canadian Patent Application No. 439,608, can give some
difficulties in commercial practice. `
Alcoholates in which the metal is titanium, aluminium
or zirconium have the advantage that they are liquids
that are miscible with the reaction mixture and so the
handling difficulties are avoided. However they are less
reactive with the result that the process must be
conducted at elevated temperatures, generally above
70C, and even at such temperatures the rate of reaction
may not be as fast as can be achieved at much lower
temperatures using, for instance, the magnesium
alcoholate. Also the use of elevated temperatures is
undesirable both for economic reasons and, in the context
of production of vinyl esters, because of the tendency to
promote the foxmation of unwanted polymers. For instance
~24~2~
-- 3 --
during prolonged operation the viscosity of the reaction
mixture may gradually increase due to the formation of
polymeric products. These may include polymerised vinyl
ester and may include polymers containing the metal,
especially when the metal includes alumînium. When the
metal of the alcoholate consists solely of titanium there
may be a tendency for inorganic cross-linked products
based on aluminium to be formed during the process.
Despite these difficulties the process described in
EP 0118639 is a very great improvement on any prior
process for conducting the ester interchange.
Somewhat similar difficulties can also be encountered
when the metal alcoholate is being used as a catalyst for
an ester interchange reaction conducted in the presence
of reactive alcohol~
We ha~e now discovered that certain novel metal
alcoholates are of particular value in ester interchange
reactions, especially for the production of vinyl esters.
These metal alcoholates contain at least two metals and
are liquid at 30C and are miscible wi*h the esters that
participate in the ester interchange reaction.
The novel alcoholates have the formula
( ~2-a Ra]x 1~1 (OR )n b R6]
wherein M is selected from Mg, Ca and Ba
M2 is selected from Ti, ~r and Al
n is 3 or 4 and is the valency of M
a is a number from 0 to 1.5
b is a number from 0 to 0.75n
x+y = 1 and each is a number from 0.005 to 0.995
and is such that the compound is liquid at 30~C
R is selected from alkyl having at least 4
carbon atoms, cycloalkyl having at least 4 carbon atoms
and aminoalkyl, and
~24t3124
-- 4 --
each group R is individually selected from C1_3 alkoxy
and blocking groups that are substantially less reactive
in the ester interchange reaction than the groups OR .
Any blocking groups will normally be a longer chain
alkoxy group or will be a hydrocarbon group. Preferred
blocking groups are C8 30 alkoxy groups and alkyl groups,
often C8 30 alkyl. The preferred blocking group is
C18 alkoxide.
The alcoholate may include two or more different
groups R6, ~or instance it may include both blocking
groups and C1 3 alkoxy groups, but the catalyst before
use is preferably free of C1 3 alkoxy groups and either
contains no groups R (a and b are both zero) or contains
blocking groups (generally a and b are each up to 1).
The catalyst generally includes only one group R since
if it contains different groups R it will result in the
production of mixed esters.
R4 is generally cycloalkyl of 5 to 8 carbon atoms,
alkyl of 5 to 30 carbon atoms or, preferably, aminoalkyl.
When R4 is a long chain alkyl it often contains 15 to 30
car~on atoms, preferably 20 to 24 carbon atoms. ~hen
R4 is a shorter chain alkyl it is often 6 to 15 and
generally 7 to 10 carbon atoms. Preferably R is an
aminoalkyl group and the alkyl group of the amino alkyl
radical generally contains at least 2 carbon atoms and
the radical is preferably a dialkyl aminoalkyl group of
the formula (R )2N-CnH2n, where n is two or three and the
groups R5, which may be the same or different, are C1 3
alkyl. Preferably the aminoalkyl group is
dimethylaminoethyl.
The values of ~ and y must be selected such that the
catalyst is li~uid at 30C. Generally x is below about
0.5 and often is below about ~.3. Generally it is above
0.01. Best results are generally obtained when x is ~rom
about 0.02 to about 0.1 or 0.2. If a catalyst is
~ :,
8~2~ _
-- 5 --
formed that is solid then it is generally necessary to
modify it by reducing the value of x. For instance when
Ml is Ca or Ba it is often desirable for x to be below
about 0.1 or 0.2.
Although the novel alcoholate has been shown above as
having an empirical formula consisting of two components,
component X being formed from the divalent metal and
component Y being formed from the trivalent or
tetravalent metal, it should be appreciated that this
representation of the alcoholate is used for simplicity
and that the precise structural relationship between the
groups is at present not entirely clear.
The alcoholate may be a simple physical mixture of
components X and Y or it may be a chemical combination.
For instance it may be a mixture of, e.g. titanium
alcoholate and magnesium alcoholate. The alcoholate can
be made by conventional methods for the production of
metal alcoholates modified by the use of a mixture of
metals, rather than a single metal. For instance an
2~ alcohol R40H may be reacted with an appropriate mixture
of metals M1 and M2 or may be reacted independently with
each of the metals and the two reaction products then
combined. Compounds of metals Ml and/or M2 may be used
in place of the metals. The preferred method of making --
the alcoholate is by an alcohol interchange reaction as a
result of reaction of an alcohol R40H with a Cl 3
alkoxide either of a mixture of the metals Ml and M2 or
of the individual metals Ml and M2 followed by
combination of the resultant alcoholate. Generally the
reaction is taken to completion as a result of which the
final product will be free of lower alkoxy groups as R6
but blocking groups may have been introduced in
conventional manner.
Each of Ml and M2 may themselves be a mixture of two
or more o the defined metals. ~1 is preferably magnesium
~Z4~3~2~ _
, .. .
-- 6 --
or calcium. The preferred alcoholates are those in which
M is magnesium and M is titanium or, less preferably,
zirconium.
The novel alcoholates may be used in ester
interchange processes either as a catalytic component or,
preferably, as a reactant that supplies alkoxy groups
during the process or as both. A process according to
the invention is one in which a compound R3CoOR~ is made
by ester interchange with a compound of the formula
R3CoORl in the presence of an alcoholate derived from an
alcohol R40H and, optionally, in the presence also of the
alcohol R OH, wherein R is Cl 3 alkyl, R is an
optionally substituted hydrocarbon, generally aliphatic
hydrocarbon, group and R is as defined above, and is
characterised in that the alcoholate is present as a
solution in the reaction mixture and is a mixed
alcoholate of formula I above.
The alcoholate is generally formed before its
introduc ~ on into the ester interchange reaction mixture
but it may be formed within the reaction mixture. For
instance either or both of the metal alcoholate
components may be formed within the reaction mixture, for
instance by reaction of alcohol R OH with metals M
and/or M or compounds thereof. It appears possible
that processes in which the mixed metal alcoholate
contributes groups oR4 to the reaction mixture can be
regarded as processes in which the magnesium, calcium or
barium component of the alcoholate contributes these
groups for ester interchange with the starting ester
R3CoOR whilst the alcoholate of titanium, zirconium or
aluminium reacts with the resultant magnesium, calcium or
barium alcoholate, typically of the formula M (OR )2~ to
reform the desired component M1(oR4)2. The invention
therefore includes any ester interchange process
con~ucted in the presence of an alcoholate of formula I
:
- ~L24~12~
-- 7 --
irrespective of whether the alcoholate is a physical
mixture of alcoholates of individual metals or whether it
is a combined alcoholate containing both metals, and
irrespective of how the combined alcoholate, or the
individual components of a physical mixture of the
alcoholates, has been provided into the reaction mixture.
The process is of particular value in the production
of vinyl esters, and in particular in processes wherein
R3 is CH2=CH- or CH2=C(CH3)-. R1 is generally methyl so
that the starting ester is preferably methyl acrylate or
methacrylate but other suitable starting esters include
ethyl acrylate and methacrylate.
The reaction is generally conducted in the absence of
water.
The reaction may be conducted in the presence of
alcohol R OH to contribute some or all of the groups OR
in the final ester and so the alcoholate of formula I may
be present in such processès solely or primarily as a
catalyst and may be present in catalytic amounts.
Preferably however the ester interchange process is
conducted in the absence of water or reactive alcohol and
the mixed metal alcoholate of formula I provides all the
groups oR4 in the final ester and this is of particular
value when R is an unsaturated group as defined above as
there is a tendency for alcohol to add across the double
bond and to form unwanted by-products. The preferred
process is therefore a process broadly as described in EP
0118639 but modified by the use of the defined alcholateO
In such processes the alcoholate may be present in
molar excess, based on the starting ester, but the
reaction is preferably conducted using an excess of the
starting ester, for example 1~0 to 10 moles of ester per
mol of alcoholatP. The reaction mixture should be free
of the starting alcohol or of any other alcohol that
could react under the prevailinq process conditions, and
~2~8~L2~a
, .
-- 8 --
is generally free of any alcohol. The reaction mixture
should be substantially anhydrous. Very small amounts of
water or alcohol may be tolerated but even these may lead
to some by-product formation.
The ester interchange processes of the invention may
be carried out simply by mixing the starting ester with
the chosen alcoholate. The reaction can be conducted at
temperatures as high as 95C but best results are
achieved at temperatures below 60C, for instance
10-50C. Generally the reaction is conducted at from
20-~0C, typically around 30C.
The reaction is allowed to proceed towards, and often
reach, equilibrium. During the process some or all of
the groups OR in the alcoholate are replaced by groups
ORl. The time required to reach final equilibrium
depends upon the metal alcoholate and the reaction
temperature but is generally 10 minutes to 2 hours.
It is normally similar to the time required to reach
e~uilibrium when the metal alcoholate is formed solely
~0 from metal M and so is often i-n the range 30 to 75
minutes or less.
The process may be conducted batchwise with the
desired ester and the resultant metal alcoholate being
separated at the end of each batch but preferably the
process is conducted continuously with the separation
being conducted continuously during the process.
A prefexred process of the invention is a cyclic
process for making vinyl ester and in which metal
alcoholate formed by ester interchange in the reaction
mixture is separated from the reaction mixture and is
reacted with excess alcohol R OH in the substantial
absence of ester R3CoORl and the resultant metal
alcoholate of formula I is recycled to the reaction
mixture. This process has the advantage of achieviny
substantially total reuse of the metal whilst avoiding
~ .~
~2~ L2~ _
g
formation of azeotropes between the starting ester and
the alcohol R OH.
The preferred process of the invention involves
carrying out the following sequential steps. In step A
the ester interchange reaction is conducted between the
metal alcoholate formed from the alcohol R40H and an
excess of the ester R COOR to form a mixture containing
the ester R3CooR4, unreacted ester R3CoORl and metal
alcoholate that has at least partially been converted to
an alcoholate formed from the alcohol RlOH. In step B
this alcoholate is separated from the reaction mixture,
the ester R3CooR4 is recovered from the reaction mixture
and the ester R3CoOR1 is recycled for use in step A. In
step C the separated alcoholate from step B is reacted~
in the substantial absence of ester R3CooR1, with excess
alcohol R OH to form alcohol RlOH and alcholate formed
from the alcohol R40H. In step D this alcoholate,
formed in step C, is separated and recycled for use in
step A and the alcohol R40H is recovered and recycled for
use in step C. The alcohol R1OH is taken off as an end
product.
Generally step B is effected by separating the
alcoholate from the esters followed by separating the
esters by fractional distillation, while step C is
effected by separating the alcoholate from the alcohols
followed by separating the alcohols by fractional
distillation.
The separation of the metal alcoholate from the
reaction mixture is generally achieved by evaporation of
the more volatile organic components from the less
volatile metal alcoholate. In order to avoid prolonged
heating, this evaporation is preferably by flash
evaporation. The separated alcoholate is a compound of
formula I. If the reaction with alcohol R~OH has been
complete the compound will be free of groups R6 that are
,
-- 3L2~3124
, .
-- 10 --
lower alkoxy groups ~although it may include blocking
groups R6~ but if the reaction has been incomplete the
recycled metal alcoholate will include some lower alkoxy
groups R .
The process of the invention has the advantages of
the processes described in EP 0118639 ~high yield of
desired product, low impurities, low usage of metal
alcoholate and easier separation procedures than prior
art processes) together with major additional advantages.
Because of the stability of the metal alkoxide and of
vinyl ester with which it is in contact it is possible to
store the regenerated mixed metal alkoxide of the
invention diluted with the starting methyl acrylate or
other ester at convenient low temperatures without
incurring the risk of metal alkoxide polymerisation. The
process can be operated at lower temperatures, as a
homogeneous system, but at reactivity rates similar to or
better than those previously attainable only in
heterogeneous systems or at much higher temperatures in
homogeneous systems. The metal alcoholate can be reused
for much longer periods than were possible in the prior
process without unacceptable build-up of by-products,
increase in difficulties of separation, or loss of
activity. The risk of polymerisation of methylacrylate
or other vinyl ester in the homogeneous reaction mixture
is reduced. The difficulties of separation and
recycling associated with prior heterogeneous systems,
and the risk of the formation of titanium based polymers
in prior homogeneous systems, is reduced or eliminated.
The following are some examplesO
EXAMPLE 1
A catalyst of formula I wherein M1 is Mg, M2 is Ti, a
and b are zero, x is 0.05 and R4 is dimethylaminoethyl is
made by reacting excess dimethylaminoethanol (DMAE~ with
a 5:95 molar mixture of magnesium diisopropoxide and
24~3~2~
-- 11 --
titanium tetra-isopropoxide. The reaction is conducted
by heating the alcohol with the isopropoxides in a
stirred reactor with a fractionation column and reflux
splitter. Isopropanol is removed at the top of the
column and vacuum is applied at the end of the reaction
to remove the last traces of isopropanol. The reactor
contains the desired alcoholate of formula I as a mobile
liquid having no sign of precipitate. Analysis shows
the product to be free of alcohol and isopropoxide groups
and to have to have the molar analysis Mg = 0.05, Ti =
0.95, dimethylaminoethoxy = 3.9, corresponding to the
formula
[Mg (DM~E)2]0 05 [Ti (DMAE)4]0.95
In a similar manner it is possible to form the
corresponding alcoholates of, for instance, calcium and
titanium, magnesium and zirconium, magnesium and
aluminium, and barium and titanium. In a similar
manner it is possible to form magnesium titanium, and
other metal, alcoholates having values of~x up to 0.5.
In a similar manner it is possible to form
alcoholates wherein R is, for instance, 2-ethylhexyl or
other higher alkyl or cycloalkyl group.
Alcoholates containing a blocking group R may be
made by starting with, for instance, titanium C18
alkoxide triisopropoxide, instead of titanium
tetraisopropoxide.
EXAMPI.E 2
300 grams of metal alcoholate is reacted with 3~4
grams of methylacrylate at varying temperatures. The
alcohol is DMAE. The molar ratio of methylacrylate to
the dimethylaminoethanol ligand (DMAE) is 4:3. The
formation o~ dimethylaminoethyl acrylate (DMAEA) is
monitored by Gas Chromatography throughout the reaction~
The results are then plotted assuming pseudo first order
kinetics and the rate constant obtained. The higher the
rate constant, the more rapid the reaction.
.~ .
. .
~Z4~.2~ _
- 12 -
The reaction is conducted twice at each temperature.
In reaction series A the metal alcoholate is Ti(DMAE~4.
In series B the metal alcoholate is the titanium
magnesium alcoholate whose analysis and manufacture is
given in Example 1. The results are shown in Table I.
TABLE I
Temperature Rate Constant
(C) ' (min-l~ x 102
A B
- 3.57
- 5.25
1.775 6.73
3.399 7.3
5.536 9.75
9.75
14.2
This shows the Mg/Ti reagent to result in a much more
rapid rate of reaction than the titanium alone at the
same temperature. The Mg/Ti compound is capable of
inducing reaction to DMAEA even at temperatures as low as
30C.
A similar clear trend is observed when comparisons
are made between Ti and TiQ 95 CaO 05, Zr and ZrO 95
MgO 05~ Al and Alo.gs MgO.os A similar trend is
observed when the alcohol is 2-ethylhexanol instead of
DMAE.
EXAMPLE 3
Using a molar ratio of 5/3 methyl acrylate to
dimethylaminoethyl ligand the rates of reaction of
Ti(DMAE)4 versus the Ti Mg alcoholate made in E~ample 1
are determined at room temperature ~20C1 by measuring
- 13 -
the percentage of DMAEA in the reaction mixture after
various reaction times in two reactions. In Reaction B
30 g of the mixed alcoholate is reacted with 43 g methyl
acrylate. In Reaction A 50 g Ti(DMAE)4 is reacted with
71.6 g methyl acrylate.
TABLE 2
Time ~mins) Reaction A Reaction B
1 ~-9
0 15.8
0.9 19.0
2.4 21.5
3.7 22.2
15 60 5.0 21.5
This shows the faster rate of reaction using
magnesium and titanium compared to the use of titanium
alone.
EXAMPLE 4
-
3:1 mole ratio mixtures of methylacrylate (MA) and
dimethylaminoethanol (DMAE) are rea~ted together at 40C
in the presence of (A) Ti(DMAE)4 and (B) Tio 5
MgO 5(DMAE)3 at a level of 2.6 mole percent catalyst
based on DMAE.
Samples of the mixtures are taken at intervals and
the mole ~ conversion of DMAE to DMAEA was recorded. The
results obtained are as follows:
. .
.
.,:`
~24~
- 14 -
TABLE 3
TIME (mins) A B
O
510 1.0 38.8
- 52.9
3.5 57.9
S0 - 63.0
6.5 66.1
10 This clearly demonstrates the advantage of catalyst
according to the invention in an ester interchange
conducted using a reactive alcohol.
EXAMPLE 5
In reaction A, 87.2 g of Zr(DMAE~4 are weighed into a
lS 3-neck round bottom flask fitted with a stirrer and
placed in a water bath held at 40C. 112.8 g MA are
added (molar ratio MA:DMAE = 5:3). The time is noted
as time zero and samples of the reaction mixture are
taken at varying times ther~ter and analysed by GLC to
determine the percentage DMAEA in the reaction mixture.
In series B the process is repeated except that the
alcoholate is zrO 95 MgO oS (DMAE)3 9- The results are
shown in Table 4.
TABLE 4
25 Reaction Time (mins) A B
S 2.5 4.5
4.9 6.8
lS 8.1 12.3
11.2 14.2
30 25 13.5 19.3
15.1 20.4
22.5 28.7
26.4 30.2
-~ ~z~x~ -
This shows that the incorporation of a small amount
of magnesium gives a significant improvement in the rate
of formation of DMAEA.
EXAMPLE 6
To demonstrate the effect of varying the proportions
of titanium and magnesium in the metal alcoholate a
series of reactions as in Example 2 are conducted ~nd the
percentage DMAEA in the reaction mixture is recorded at
5, 30, 60 and above 180 minutes from the start of the
reaction. The results are shown in Table 5.
T~BLE 5
Reaction100 Ti 95 Ti 90 Ti 80 Ti 50 Ti 0 Ti
Time 0 Mg 5 Mg 10 Mg 20 Mg 50 Mg 100 Mg
155 0 2.2 3.8 7,5 (16)7,7
2.4 (10~ 9.1 14.3 24.519.5
5.0 17.2 (12) 16.8 25.1(23)
>180 2~.0 26.5 25.2 27.8 25 27.5
~ The values in parent ~ sis are obtained py interpolation.
EXAMPLE 7
5 moles methylacrylate are reacted with 1 mole of the
titanium magnesium alcoholate made and analysed in
Example 1. The reactants are charged to a reaction
vessel maintained at about 40C. Equilibrium is obtained
after about 60 minutes and the volatiles ~including
DMAEA) are removed by xapid distillation at 120C and 20
mm Hg leaving a liquid residue. The DMAEA that is
separated from the volatiles is found to be substantially
pure. The liquid residue is regenerated with 1 to 2
moles DMAE per mole of methoxide in the residue and
methanol and DMAE are removed slowly by distillation on a
rotary evaporator. The residue is recycled for further
reaction with methylacrylate.